Motion processing

Motion processing behind occluders

Luca Battaglini, Gianluca Campana and Clara Casco

Can our brain compute motion when a moving target is occluded by another object as a tunnel? De Lucia & Liddel 1998 suggested two different hypothesis: 1) estimation of occlusion time and counting to predict the time to contact (TTC) between the target and the end of the tunnel, 2)extrapolate the motion behind the occluder maintaining active the speed information. Observers compared either the speed (when non occluded) or the TTC (when occluded) of moving stimuli differing in contrast or size. Our results showed that, in agreement with the literature, smaller or higher contrast stimuli were perceived as moving faster, but the major result showed that the TTC was shorter with these stimuli. This indicates that the perceived speed influences the elaboration of motion once occluded, thus supporting the perceptual hypothesis.

Temporal characteristics of perceived non-overlap of objects in audiovisual stream/bounce displays

Yousuke Kawachi and Yoshihiko Matsue

In a stream/bounce display in which two identical visual objects move toward each other, coincide (overlap), and then move apart, the objects can be perceived as either streaming through or bouncing off each other. Our previous report (Kawachi, 2011, i-Perception, 2(8), 834) demonstrated a perceptual phenomenon in which the presentation of a brief tone at the coincidence point promoted the perception of a non-overlap and of the objects bouncing off each other.. In the present study, the amount of overlap between two objects and the temporal offset between the overlap and the tone presentation (0, ±90 ms, and ±390 ms) were parametrically manipulated. Observers were first asked to judge whether the two objects overlapped with each other and then asked whether the objects appeared to stream through or bounce off each other. When the tone was presented at and before the overlap, the perception of a non-overlap and of the objects bouncing off each other was stronger than when the tone was presented after the overlap. Together with the results of a supplementary experiment on task-relevant selective attention to a visual event, we discuss how an abrupt auditory stimulus temporally modulates visual processing in a stream/bounce display.

Anisotropy in postural sway caused by optic-flow: a matter of co-contraction?

Vivian Holten, Maarten J. van der Smagt, Stella F. Donker and Frans A.J. Verstraten

Previous research showed that contracting radial optic-flow patterns generate more sway than expanding patterns [Wei et al, 2010, Journal of Vision, 10(14):4, 1-10]. The origin of this (expansion vs. contraction) anisotropy is still unknown. Since expanding patterns are more commonly experienced when moving through the environment, the lesser sway for this condition might be related to anticipatory co-contraction of antagonist lower leg muscles. To investigate this, we presented radial random-dot optic-flow patterns to participants, standing on a force-plate in a completely dark room. Simultaneously, electromyography (EMG) from the tibialis anterior and the gastrocnemius was recorded. Optic-flow stimuli (duration 4s) were interleaved by dynamic visual noise patterns (duration varied between 3.4-4.2s). Postural sway was measured in the anterior-posterior direction. While our results concur with the earlier reported anisotropy in postural sway, we do not observe co-contraction during expanding (or contracting) optic-flow pattern stimulation. Instead, lower-leg muscle activity appears to just follow sway direction, indicating that the (expanding vs. contracting) anisotropy in optic-flow induced postural sway is not caused by anticipatory co-contraction of antagonist lower leg muscles.

Visual traveled distance estimation by leaky integration in driving situations

Maren Stiels, Harald Frenz, Jochen Müsseler and Markus Lappe

The perception of travel distance in walking range shows characteristic underestimation that can be explained by leaky path integration with gain and leak rate as parameters. We wanted to test whether this is true also in driving, in which distances are much longer and velocities much higher. Thirty subjects drove a multimodal, interactive driving simulator that consisted of a mock car, a 2.80 x 1.60m projection screen, shakers at the car carriage, and vehicle acoustics. In the simulations, subjects drove at speeds up to 160km/h along a straight road with surrounding trees. The experimenter stopped the simulation at specified distances between 500 to 5000m and subjects reported their travel distance estimate in meters. In a second condition, subjects were given a distance number (e.g. 3500m) and were asked to drive and stop when they felt they reached that distance. The data from both conditions were fit with the leaky integration model. We found participants able to discriminate even long travel distances quite accurately, but with a difference between the two conditions that is characteristic for the leaky integration model. The gain parameter was higher than for walking, whereas the leak rate was smaller for driving than for walking.

Factors that affect tracking of occluded target motion

Anna Hughes, Martin Stevens and David Tolhurst

In natural scenes, a moving object is very often occluded from view at various points along its trajectory, and it has been shown that humans are able to track invisible target motion created by occlusion. We examined the accuracy of this occlusion tracking using Gabor patches which moved laterally across a CRT before disappearing at the edge of the display; we investigated the effect of the orientation and motion of the stripes within the Gabor. The pattern orientation of the Gabor patch affected the speed judgements, with oblique stripes perceived as moving faster on average compared to vertical or horizontal stripes. Introducing pattern motion within the Gabor patch also caused systematic changes in perceived speed. These changes were in accordance with whether the pattern appeared to be moving forwards or backwards relative to the overall direction of motion; the patches were judged to move faster than a 'no motion'control when the pattern movement was forwards and were judged to move more slowly than a 'no motion'control when the pattern movement was backwards. The findings are discussed in terms of animal camouflage mechanisms and how this research can aid understanding of motion perception in the human visual system.

Interaction between fine and coarse scales impairs motion discrimination more strongly than surround suppression

Ignacio Serrano-Pedraza, Maria J. Gamonoso, Vicente Sierra-Vázquez and Andrew M. Derrington

Ability to discriminate motion direction of a Gabor patch diminishes with increasing size and contrast, indicating surround suppression. Discrimination is also impaired by a static low-spatial frequency patch added to the moving stimulus, suggesting an antagonism between sensors tuned to fine and coarse features. Using Bayesian staircases, we measured duration thresholds in motion-direction discrimination tasks using vertically oriented Gabor patches moving at 2deg/sec. We tested two contrasts (2.8% and 46%) and five window sizes (from 0.7 to 5deg) and two spatial frequencies, 1c/deg and 3 c/deg, presented either alone or added to a static component. When the moving component was presented alone, duration thresholds increased with size at high contrast, and reduced with size at low contrast. When a 1c/deg moving component was added to a 3 c/deg static component duration thresholds where similar to the case when the moving component was presented alone. However, when a moving 3c/deg component was added to a static 1c/deg component duration thresholds increased about four times for high contrast and two times for low contrast. These results suggest that the effect of surround suppression is small compared with the effect produced by the antagonism between sensors tuned to fine and coarse features. [Supported by Grant No. PSI2011-24491 from Ministerio de Ciencia e Innovaci ón (Spain)]

Development of Optic Flow Sensitivity in School-aged Children of Kathmandu

Mahesh Raj Joshi and Helle K Falkenberg

The development of sensitivity to optic flow patterns was investigated by measuring motion coherence thresholds (MCTs) in school-aged children in Kathmandu. 119 child observers aged 6-16 years and 24 adult observers (mean age, 23.66 years) participated with informed consent. The observers identified the direction of a translational (rightward vs leftward) and radial (expanding vs contracting) optic flow pattern with 100 dots of 75% (Michelson's contrast) moving at 5.48°/s and 1.56°/s in a 2AFC task. The direction of each dot was drawn from a Gaussian distribution whose standard deviation was either low (similar directions) or high (different directions). MCTs were measured separately for each speed and optic flow pattern. The sensitivity to radial optic flow improved with age, and were still immature at 16 years (p<0.05) at both speeds. All observers had better sensitivity at 5.48°/s compared to 1.56°/s (p<0.05), with difference in developmental pattern for children at these speed. Sensitivity to radial optic flow is immature until late childhood. Differences in sensitivity and development of radial optic flow at two speeds, suggest different motion channels involved in processing slow and fast speeds.

Apparent duration is compressed for accelerating visual motion

Alan Johnston, Aurelio Bruno and Inci Ayhan

Johnston (In Nobre and Coull (eds), Attention and Time, 2010) proposed a 'predict and compare' clock to explain the effects of adaptive state on the perceived duration of a drifting stimulus. The clock accumulates the number of satisfied (constant velocity) temporal predictions. We might expect an interval containing acceleration should appear expanded, since the stimulus always preempts the prediction. Here we studied the effects of acceleration and deceleration on the perceived duration of a drifting grating. We used several velocity ranges (minimum 0; maximum 67 deg2/s) over three standard durations (300, 600 and 900 ms). The average speed was kept constant (10 °/s). We found that increasing the velocity progressively reduced perceived duration. However the reduction was proportional to range (independent of standard duration) rather than acceleration. Also the reduction was proportional to the standard length indicating a constant effect through the interval. The results could not be explained by changes in apparent onset or offset. The effects of increases and decreases in speed were highly asymmetric. A decelerating stimulus was subject to a mild dilation that was more pronounced for low rates of deceleration. We suggest acceleration may differentially activate high temporal frequency channels leading to duration compression.

Alternation rate dependency of perceptual asynchrony in color-motion binding

Osamu Watanabe, Kazuhiro Kanahama and Yukinori Suzuki

It has been demonstrated that psychophysical judgments on temporal synchrony of alternating color and motion direction show an apparent delay of motion perception relative to color perception [Moutoussis and Zeki, 1997, Proceedings of the Royal Society of London B, 264, 393-399]. Recent researches showed that the perceptual asynchrony depended on experimental tasks and the alternation rates, and these findings imply that the phenomenon does not simply reflect neural delays of color and motion processes [Nishida and Johnston, 2002, Current Biology, 12, 359-368]. Here we examine what causes the alternation rate dependency of the perceptual asynchrony measured by the pairing task [Clifford et al, 2003, Vision Research, 43, 2245-2253]. To complete the task, observers had to judge durations of individual motions while a stimulus had a certain color, and the apparent delay would be due to a 'duration distortion' effect [Kanai and Watanabe, 2006, Perception & Psychophysics, 68, 1113-1123]. The results showed that the alternation rate dependency could not be fully explained by the duration distortion effect and disappeared when another motion signal indicated the periods for judgments. These results suggest that the alternation rate dependency in the pairing task is due to the temporal binding mechanism of different visual features.

The influence of concomitant pitch perception on timing of a coincidence-anticipation timing task

Toshinori Hanai and Ataru Yoshino

We examined whether the influence of concomitant pitch perception on timing of a coincidence-anticipation timing task (CAT). A white dot on black background was moved at uniform velocity horizontally from left to right across the center of a 19-inch monitor. The moving speed of the dot was 2 deg / s. The moving dot disappeared between 180 deg and 360 deg from the start of the move. It appeared after the participant's key touch. Vertical lines were displayed in the appearance point and disappearance point. In synchronization with the motion, a sound was played. We presented 16 acoustic conditions. We used 4x4, ie 16 conditions of the animated object moving with all permutations of pairs of the tones 220 Hz, 440 Hz, 880 Hz and silence exchanging after 3 s. The participants were allowed to describe their impressions freely, and they also judged the timing that dot after the occlusion was considered to reach the vertical line by the key touch. As a result, it was observed that concomitant pitch perception increased the error on timing of CAT. It was suggested that oncomitant pitch perception affect the timing task of CAT.